What is the most engaging way to survey a large audience of friends?
September 30, 2020 10:04 PM Subscribe
Have you ever taken a survey or poll that was fun / easy / rewarding / delightful?
I'm trying to figure out the most engaging way to survey a large group of friends (N=30 to 1000) on an ongoing basis.
The ideal solution:
The ideal solution:
- is easy / fun so people enjoy contributing
- gives each user a unique ID so they don't have to type their name every time
- is rewarding (e.g., provides feedback / insights, like VIEW RESULTS in a poll)
- The software organizes that data, so that I can share the results with folks who participated.
- I notice that some participants are really great at suggesting nonfiction books, so I want to be able to tag them as such.
- Survey software (Typeform, SurveyMonkey)
- Email software (Mailchimp)
- Something like a loyalty rewards program (CommunityRewards)
- Something else?
The only thing people fill out voluntarily are facebook quizzes. The topics are wildly different every time, you get to compare yourself to your friends on "which kitchen implement are you" and "are you a genius?" type questions, and nobody has to join in, but social media peer pressure sells it.
Is that even an option with your content?
posted by Omnomnom at 4:04 AM on October 1, 2020 [2 favorites]
Is that even an option with your content?
posted by Omnomnom at 4:04 AM on October 1, 2020 [2 favorites]
Best answer: Part of my job involves analyzing survey data for anywhere between 50-10k people, using either Redcap (terrible interface) or Qualtrics (slightly less terrible interface). I have never met or designed a "fun" survey, unless you count those Buzzfeed "Which Jane Austen character are you?" quizzes that were all the rage a few years ago. (On preview, what Omnomnom says.)
The #1 rule is shorter is better. Responses drop off considerably once you get past 3-5 questions. I once offered a $1000 raffle as an incentive for a "long" (estimated 10 minute) but important survey. Despite that, only around 20% started it, and still about 1/3 of respondents still abandoned it halfway through.
Step 1 is to decide what's important to you and to have that upfront. For your book recommendation example, that would probably be free-text fields for Title and Author. Maybe an optional field to allow people to say why they are recommending that book. Note that book recommendations are going to be inherently fluid; the book I'd recommend to my friend when they are feeling upbeat is going to be different from a book I'd recommend after they just found out their spouse is having an affair.
Note also that this format, while easy and minimally unpleasant for the respondent, is a nightmare for you to deal with. People are going to put all kinds of crap in free-text fields (misspelled words, non-words, etc) which you will have to "fix" before presenting the data. From your perspective, a poll where people select or rank their preferred choice out of 5-10 is easiest FOR YOU. A poll also allows people to view results pretty much immediately. But then you have to pre-select the 5-10 choices ahead of time, from a field of nearly infinity options, and frame it as "what should I read next?" rather than "what's the best book on this list"
Other thoughts: Is longitudinal data (unique ID) important? Why? How are you planning to judge the "best non-fiction reader?" Does anyone care to have that tag? Will that inhibit responses from other users?
posted by basalganglia at 4:25 AM on October 1, 2020 [2 favorites]
The #1 rule is shorter is better. Responses drop off considerably once you get past 3-5 questions. I once offered a $1000 raffle as an incentive for a "long" (estimated 10 minute) but important survey. Despite that, only around 20% started it, and still about 1/3 of respondents still abandoned it halfway through.
Step 1 is to decide what's important to you and to have that upfront. For your book recommendation example, that would probably be free-text fields for Title and Author. Maybe an optional field to allow people to say why they are recommending that book. Note that book recommendations are going to be inherently fluid; the book I'd recommend to my friend when they are feeling upbeat is going to be different from a book I'd recommend after they just found out their spouse is having an affair.
Note also that this format, while easy and minimally unpleasant for the respondent, is a nightmare for you to deal with. People are going to put all kinds of crap in free-text fields (misspelled words, non-words, etc) which you will have to "fix" before presenting the data. From your perspective, a poll where people select or rank their preferred choice out of 5-10 is easiest FOR YOU. A poll also allows people to view results pretty much immediately. But then you have to pre-select the 5-10 choices ahead of time, from a field of nearly infinity options, and frame it as "what should I read next?" rather than "what's the best book on this list"
Other thoughts: Is longitudinal data (unique ID) important? Why? How are you planning to judge the "best non-fiction reader?" Does anyone care to have that tag? Will that inhibit responses from other users?
posted by basalganglia at 4:25 AM on October 1, 2020 [2 favorites]
I LOVE SURVEYS and I would be happy to be one of your surveyees!
I think the key to making them fun is to make them easy to do--the more you can click buttons and the less you have to write free text, the better.
posted by exceptinsects at 12:14 PM on October 1, 2020 [3 favorites]
I think the key to making them fun is to make them easy to do--the more you can click buttons and the less you have to write free text, the better.
posted by exceptinsects at 12:14 PM on October 1, 2020 [3 favorites]
Make sure people know why it's important.
Don't harvest data that's too personal. I bail on surveys that feel invasive.
Buzzfeed-style assessments. What kind of Great Ape are you? Babboon, Gorilla, Bonobo, orangutan, etc. And funny, like, You're a bonobo, distinguished by relatively long legs, pink lips, parted long hair on its head, excellent sense of humor, mathematical talent and love of bowling. Only, you know, funnier, and with links to pictures and .gifs.
I like surveys and books, so let me know if you need participants.
posted by theora55 at 1:12 PM on October 1, 2020 [2 favorites]
Don't harvest data that's too personal. I bail on surveys that feel invasive.
Buzzfeed-style assessments. What kind of Great Ape are you? Babboon, Gorilla, Bonobo, orangutan, etc. And funny, like, You're a bonobo, distinguished by relatively long legs, pink lips, parted long hair on its head, excellent sense of humor, mathematical talent and love of bowling. Only, you know, funnier, and with links to pictures and .gifs.
I like surveys and books, so let me know if you need participants.
posted by theora55 at 1:12 PM on October 1, 2020 [2 favorites]
In my eyes, this question is missing the bigger context of how you’re reaching and managing your audience.
Instagram stories are quick, flexible, integrated with social media, has a variety of question styles. Can be shareable or locked to your followers. People already have accounts.
You can share results back to your stories feed, archived as a highlight, or on your grid.
There is a graph api for accessing data, but I don’t have a business/creator account to give you any better info about it, but a light googling told me you may be able to extract poll data that way.
The account @notestostrangers has a large following and from time to time will do a “suggest a poll” story (open ended question format) and post the replies (in stories) as an A/B question. Sometimes these go up as a dozen or more at a time. You see current results as you respond.
posted by itesser at 6:48 PM on October 1, 2020 [1 favorite]
Instagram stories are quick, flexible, integrated with social media, has a variety of question styles. Can be shareable or locked to your followers. People already have accounts.
You can share results back to your stories feed, archived as a highlight, or on your grid.
There is a graph api for accessing data, but I don’t have a business/creator account to give you any better info about it, but a light googling told me you may be able to extract poll data that way.
The account @notestostrangers has a large following and from time to time will do a “suggest a poll” story (open ended question format) and post the replies (in stories) as an A/B question. Sometimes these go up as a dozen or more at a time. You see current results as you respond.
posted by itesser at 6:48 PM on October 1, 2020 [1 favorite]
Yeah, the only area I've seen people doing this in a way that's fun and not a chore is via Instagram or Twitter polls. Twitter polls aren't as broad as a survey (just one question per tweet and I think only 4 possible choices) but you get the immediate feedback of the results, and there are a bunch of accounts who are doing like, movie brackets and stuff via these polls, so you can root for your favorites/bemoan the idiots who didn't vote for your favorite, etc. On Instagram, I see stuff more like your example of looking for book recommendations - people post as a "question" prompt asking for recommendations of books/food/makeup/etc, and then they can screenshot and share all the responses they got so their other followers also benefit from the answers. I imagine there's similar functionality on Facebook although I'm less active there. So you might start by just posting polls/questions on social media that you already have (and presumably are already connected with your friends on) before you try to use a whole separate survey tool, and see if that scratches your itch!
posted by jouir at 12:17 PM on October 2, 2020 [1 favorite]
posted by jouir at 12:17 PM on October 2, 2020 [1 favorite]
This thread is closed to new comments.
No. Not even once. Surveys are a pain in the arse, and if I fill one in it's pretty much always because doing so is either the least bad option or a sacrifice of time I'm willing to make to serve a worthwhile larger purpose.
I'm trying to figure out the most engaging way to survey a large group of friends (N=30 to 1000) on an ongoing basis.
Nobody has 1000 friends, and becoming known as the person responsible for choosing to survey 1000 people on an ongoing basis will pretty much guarantee the truth of that in perpetuity.
posted by flabdablet at 3:38 AM on October 1, 2020 [17 favorites]